Having a partner’s worn shirt nearby may feel reassuring at bedtime. One study tested this familiar experience using wrist-worn activity monitors and questions each morning. In data from 155 people, sleep efficiency was higher on nights when a partner’s shirt was placed over the pillow than on nights using a control shirt.
Interestingly, instrument-estimated sleep and people’s sense of having slept well did not follow exactly the same pattern. The 2020 study considered both the scent of a close partner itself and the belief that “this is my partner’s scent.”
Four nights sleeping alone, with a change of shirts
The researchers recruited opposite-sex couples who had been together for at least three months. Sleep data were analyzed for 155 people across three samples, with a mean age of about 21. Participants were screened to exclude conditions such as sleep disorders and chronic illnesses.
The first sample comprised 37 women, the second 40 women and 38 men, and the third 40 women. Each person slept for two nights with their partner’s shirt and two nights with a control shirt. The control was an unworn shirt in the first two samples and a stranger’s worn shirt in the third. This was a pooled analysis of three related designs, not a study of 155 couples under one identical condition.
Scent donors wore a cotton T-shirt for 24 hours while using unscented washing products and avoiding scented products and activities likely to affect odor. The shirts were frozen after collection. Sleep participants received two visually identical shirts, using one over their pillow on Monday and Tuesday and switching to the other on Wednesday and Thursday. The order was randomized, and both participants and the staff working with them were kept unaware of the conditions.
Participants slept alone on all four nights. Procedures were also standardized, including washing bedding with unscented detergent when changing shirts. These conditions differ from a comparison involving actually sharing a bed with a partner.
Sleep efficiency rose by about two percentage points on average
The wrist-worn activity monitor estimates sleep and wakefulness from bodily movement. “Sleep efficiency” in this study was the proportion of time spent in bed trying to sleep that was estimated to have been spent asleep. It distinguishes simply staying in bed longer from sleeping more continuously while there.
The mean was 88.03% on partner-shirt nights and 85.35% on control nights. A statistical model accounting for repeated measurements also found a significant difference of about 2.58 percentage points associated with the partner’s shirt. The main conclusion was unchanged after accounting for factors such as daily stress, relationship duration and quality, and presentation order.
An internal meta-analysis that treated the three samples separately before combining their results found a mean difference of 2.14 percentage points, with a 95% confidence interval of 0.61–3.68 points. The summary “about two percentage points” refers to this pooled result. It combines three samples within the same research project, not replications conducted by separate research teams.
Because the measure is estimated from actigraphy, it does not establish that deep sleep or any particular sleep stage increased. Still, it is meaningful that the study detected a difference based on movements during sleep rather than relying only on questions about how people felt.
Perceived sleep quality also involved recognizing the scent
Each morning, participants rated overall sleep quality and how rested they felt on seven-point scales. The mean was 4.74 with the partner’s shirt and 4.60 with the control, but the initial analysis did not reach statistical significance. In an analysis adjusting for that day’s stress and the day of the week, the partner-shirt effect became significant.
Subjective sleep therefore cannot simply be described as “unchanged.” The difference was small enough that the conclusion depended on how daily circumstances were taken into account.
The researchers also examined whether participants believed the shirt belonged to their partner. Overall, they identified the condition correctly about 70% of the time. When actual shirt type and participants’ beliefs were included in the same model, sleep efficiency was associated with actually using the partner’s shirt, while the effect of believing it was the partner’s was not clear.
By contrast, subjective sleep quality was significantly associated with believing the scent belonged to the partner, and the effect of the actual shirt type was no longer clear. Instrument-estimated sleep efficiency and perceived sleep quality were only weakly correlated. The findings show the value of both measures for understanding responses to a familiar scent.
How far can the effects of a close person’s scent be explained?
The authors considered the possibility that a partner’s scent evokes their presence, promoting a sense of security or reducing vigilance. However, this study did not directly establish those psychological or physiological pathways. Nor did it randomly manipulate beliefs about the shirt’s identity, so it does not isolate a causal effect of belief alone.
The difference was smaller in the sample whose control was a stranger’s shirt than in the samples using an unworn shirt. This offers a clue that the control scent might also have some effect, but the design alone cannot establish that a stranger’s scent improves sleep.
Participants were mainly young and healthy, and observations covered four nights. Persistence over longer use and effects in people with insomnia remain questions for future research. Although the original paper compares effect sizes with existing melatonin research, it was not a trial comparing medication and shirts in the same participants.
Summary
On nights when a partner’s worn shirt was placed over the pillow, actigraphy-estimated sleep efficiency was higher than with an unworn shirt or a stranger’s shirt. Across the three samples, the pooled difference averaged about 2.1 percentage points.
Meanwhile, the feeling of having slept well was also related to daily stress and recognizing the scent as a partner’s. A familiar person’s scent may relate not only to the experience of sleep but also to what happens during sleep. Examining these separately is an important finding of this study.
The research received support from the American Psychological Foundation, the University of British Columbia, the Natural Sciences and Engineering Research Council of Canada, and other sources.
Reference
Hofer MK, Chen FS. The Scent of a Good Night’s Sleep: Olfactory Cues of a Romantic Partner Improve Sleep Efficiency. Psychological Science. 2020. doi:10.1177/0956797620905615
